Predictive coding apparatus

摘要

A predictive coding apparatus encodes an input video signal. A prediction error signal of the input video signal is generated by at least another video signal which is coded before or after the input video signal. The prediction error signal is translated into codes of a fixed length with a predetermined quantizing step width. Activity of the prediction error signal is detected and compared with a reference value which depends on the quantizing step width. The codes of the fixed length or predetermined codes of zero values are selected per pixel block in accordance with the comparison result, the selected codes being used as the coded other video signal. The selected codes are translated into codes of variable lengths which are outputted as a coded video signal. Detected instead of the activity may be the maximum value and the number of non-zero values per pixel block of the codes of the fixed length or the maximum value per pixel block and the sum of the maximum values in pixel blocks of the codes of the fixed length which are compared with different reference values.